top of page
Learned Gaussian Quadrature

  • Developing a machine learning-based quadrature rule – the learned Gaussian quadrature (L-GQ) 

  • L-GQ significantly improves the computational efficiency for constructing the system matrix.

  • L-GQ can be readily applied to various modern FE formulations, such as enriched FEs, meshless schemes, and continuum-based 3D beam FE. 

ML-based Accelerated & Inverse Analysis for Focused Ultrasound

  • Determining the intensity, focal position, and shape of ultrasound remains challenging task, and computational simulations are not suitable for clinical practice due to high computational costs.

  • ML-based  inverse and accelerated analysis techniques allows for real-time navigation of transducers and estimation of acoustic fields.

bottom of page